Here are all the posts in this blog, latest on top. Bold titles are the most popular. Consider starting with Why this blog and consulting the terminology page. Also, the topics page describes all the topics (and tags), and can serve as a high-level map.
2021:
- About “The coming AI hackers” paper
- ADS residual risk – some related research
- Estimating the residual risk of ADAS / AV
2020:
2019:
- M-SDL, the autonomous vehicles verification language, is now open
- AV coverage and performance metrics
- Misc. stuff: ASAM, DeepMind, Tesla and more
- Don’t stay in Monte Carlo (for AV verification)
2018:
- Bridging AV verification and AV regulation
- Misc stuff: A Michigan conference, and some overviews
- Stuttgart impressions: Scenarios and problems
- What I have been doing lately
- The Uber accident and the bigger picture
- How to write AV scenarios (and some notes about Pegasus)
2017:
- Verifying how AVs behave during accidents
- On Mobileye’s formal model of AV safety
- Using program induction for verification
- AV verification updates: Waymo, Apple and more
- What’s new in AV verification: Stuttgart report part two
- Where Machine Learning meets rule-based verification
- Dynamic verification in one picture
- What’s new in AV verification: Report from the Stuttgart symposium
- DeepXplore and new ideas for verifying ML systems
- Topics in this blog
- One-shot imitation learning and verification
- Some papers about AV verification
- Misc stuff: Mobileye, simulations and test tracks
- Misc stuff: The Oscars bug
- Formal verification of really-complex systems
- Verifying interactions between AVs and people
- The “Synthetic Sensor Input” problem in AV verification
2016:
- Verification, coverage and maximization: The big picture
- HVC’2016 trip report
- Misc stuff: The verification gap, ML training and more
- Verification implications of the new US AV policy
- Using Machine Learning to verify Machine Learning?
- Machine Learning for Coverage Maximization
- Machine Learning verification and Explainable AI
- About faults and bugs
- Checking probabilistic components
- Future of verification: Better ways to predict behavior
- The Tesla crash, Tsunamis and spec errors
- The ICRA 2016 verification workshop
- Misc stuff: Automation and unemployment
- Finding bugs in autonomous vehicles: My impressions from the second Stuttgart symposium
- Automation and unemployment are coming, but I guess we’ll manage
- The rise of mostly-autonomous systems
- Impressions from the robotics world
- What I have been doing lately
- Meanwhile, on the Autonomous Vehicles front
- The bigger picture of verifying autonomous robots
2015:
- Misc stuff: Robotics, system simulations, AI
- HVC’2015 trip report
- Is a Universal Verification Framework possible?
- Misc stuff: HVC, game design languages and more
- Verification challenges of autonomous systems
- Security research is pretty relevant for us
- Misc stuff: HN, machine learning and Jupyter
- Terminology for multiple verification tribes
- The soup of systems
- It’s the spec bugs that kill you
- Terminology related to spec bugs
- Some good comments I got lately about requirements
- FV has much better PR than CDV
- On that “verification range” issue
- A breakfast discussion about simulations for transportation policy decision making
- Some feedback on my Stuttgart report
- Also from Stuttgart: Using simulation for public policy decision making
- My impressions from the Stuttgart Autonomous Vehicles Test & Development symposium
- Why this blog